Video cameras such as webcams used for online video communications typically have insufficient dynamic range for many scenes, which may result in objects of interest being either under or over exposed. For cameras used in video conferencing applications, the most important objects of interest are people. For efficient capture of participants in a video conference, panoramic cameras may be used.
Many cameras provide a feature called backlighting which instructs the camera's auto exposure/gain algorithm to use a predefined image region—typically the center region or an outline of a person—to get a better image of people. However, in a panoramic camera there are multiple people in the scene and standard backlighting may not work in these scenarios. In addition, face detection has been used to dynamically change the backlighting region of interest, but the image has to be good enough (image quality and facial resolution) for face recognition to provide useful results.